Reciprocating compressor flowrate target vs calculated 3

[pyMorty]

Hi everyone,

In reciprocating compressors, what does it happen when the calculated flowrate is actually higher that the source is able to provide? Let´s say the performance states the compressors would compress, under the given conditions, a flow of 10 MMscfd, but the source is just able to supply 8 MMscfd. To me it is not as simple as the compressor just handling the 8 MMScfd, due to the cylinders bore size trying to fit the flow they are physically capable to handle given the suction pressure, and the system would try to find a balance, what this balance would entail?


Thanks,
Diego

 

[albertopro]

Hi Diego,
if your compressor is cylinder type often there is the possibility to disengage one cylinder or a group of cylinders in order to reduce the machinery capacity, otherwise the pressure falls down.

 

[zdas04]

A positive displacement compressor moves the same amount of space each revolution. That is the key to keep in mind. It is moving constant physical volume, but the mass flow rate is anything but constant. If I try to pull 10 MMSCF/day from a source that can only provide 8 MMSCF/day then the suction pressure has to go down until the supplied mass fills the fixed volume. Generally the first place you see a problem is in interstage and final discharge temperatures (which are just a function of compression ratios and gas composition) and not in compressor hp which goes up with increasing ratios but goes down with reduced mass)

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[pyMorty]

Thank you Alberto and David. It is pretty much clear now that the system will compensate with gettign the suction pressure down. I'll simulate with lower pressures until the source flow is reached and see if volumetric efficiency and maybe blank off occur.

Alberto, to clarify on the capacity control, ath this point we already deactivated first stage in order to handle less flow, however the capacity is still above the source is able to provide. Unfortunatelly is the only unit available, we have to check the load of the unit, though, for it not being that low.


Thanks again.

 

[JJPellin]

Some reciprocating compressors would have another potential problem in this situation. If the discharge of the compressor is a system that holds a constant pressure, then the differential pressure across the compressor would increase. For some of our compressors, it is possible to exceed piston rod loading limits in this situation. You could overload a piston rod and have a rod failure. For some compressors, it would be impossible to exceed the rod load limit. But, you should consult with the compressor manufacturer to be sure you will not be at risk of a catastrophic rod failure. We would normally address this risk with appropriate alarms on calculated rod load or simple cylinder differential pressure. The best way to stay out of trouble is to have spill-back systems that can spill back to the suction to prevent the suction pressure from going too low.

Johnny Pellin

 

[zdas04]

DiegoGDD,
The concept of "blank off" only has meaning in dynamic compressors where a high discharge pressure can actually prevent the compressor from pushing the gas into the downstream pressure. In recips, the machine will try to push the gas until something comes apart (the "rod load" that Mr. Pellin is talking about). Generally we hit high temperature kills with a decent safety margin on allowable rod load, but these instruments fail (or get disabled by operators who don't understand the risks).

If you've crippled the first stage (it sounds like this is a 2-stage, but that is just a guess), then the only remaining alternatives are to slow the compressor down (preferred, but not all drivers can accommodate varying speed) or use a recirc/spill-back valve to dump some of the discharge pressure back to make-up missing suction flow. Recirculating gas has pretty poor operational efficiency, but it is a lot better than breaking a machine.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[pyMorty]

Thanks JJPellin and zdas04 (the unit is a 3-stage). I'm using in this case the Ariel software, I will then simulate the unit varying the suction down until it reaches the target flow, so that I can foresee the pressure it would get down to and analize possible rod load and VE issues.

I appreciate your support guys.

 

[LittleInch]

Couldn't you run the machine at 10 mmscfd with correct pressure inputs, but just re-cycle 2mmsfd from outlet to inlet? A bit of an efficiency loss alright, but keeps it in the same usage and loads, or is that too simplistic.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[pyMorty]

Thank you LittleInch. But if the source provides no more than 8 mmscfd there is no chance to recycle extra flow. The unit´s characteristics would be what state it could handle 10 mmscfd, if available, since it is not available then suction has to be adjusted for the cylinders to receive 8 mmscfd.

 

[zdas04]

DiegoGDD,
I don't think you understand what you are saying very well. At a given suction pressure and rpm, 10 MMSCF/d will go through the compressor The machine doesn't have any way to differentiate "new" gas from "recycled" gas. Often times just taking 2 MMSCF/day off the discharge and sending it back to the suction is the best short term solution. It increases your power consumption by about 25%, but it generally requires minimal capital. It is almost always possible in the short term (sometimes it requires a jumper hose from a blowdown valve on the discharge back to a blowdown valve on the suction, and hoses can be problematic in an HazOP), other times you can crack open a bypass (not very elegant and difficult to control, but possible), the best answer is an automated valve.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[gwalkerb]

To expand a little bit more on what zdas04 & LittleInch said: there are a number of ways to control the capacity of a given compressor, but the two most common are gas recycle and driver speed.

What is driving this compressor? If it's a natural gas engine you can typically turn it down to 2/3 - 3/4 RPM, or if you're using an electric motor on a VFD, you can turn that down quite a bit. The Ariel performance software will show you the adjusted flow based on RPM. If it's a single-speed electric you might be out of luck.

Secondly, a recycle valve is included on pretty much every single reciprocating compressor, at the very least for startup. Assuming you have an automated control valve in your recycle, it's probably already configured to control your suction pressure. Here's what happens: your suction pressure starts dropping, as the compressor is taking more gas than you can supply. The recycle valve will automatically open, to recycle some of the high pressure discharge gas back to suction until your suction pressure returns to it's setpoint, and the compressor then sees 10 MMSCFD. That extra 2 MMSCFD then goes through your compressor, and when it gets to the discharge, it can then recycle again through the compressor.

Think about it this way: when it first starts up, the compressor steals 2 MMSCFD from the discharge when it first starts, but then continuously recompresses and recycles that 2 MMSCFD along with the new 8 MMSCFD. (I know it's not the same gas molecules going through the recycle, but it can help to think about it this way). You do waste some energy recompressing the same gas repeatedly, but if you have an existing machine, this is the easiest way to control your capacity.

 

[pyMorty]

I got you guys, thanks zdas04 and gwalkerb for clarifying it for me.

 

[pyMorty]

Guys, one more doubt, in this case let´s say I have a separator which is feeding the compressor suction, this separator is operating at 600 psig, that is giving the 8 MMscfd, then if a recycle to get 10 MMscfd at compressor suction it will be by means of increasing the pressure to maybe 700 psig, so at the very compressor suction line the pressure will be 700 psig, whereas the separation pressure remains at 600 psig, right?

 

[LittleInch]

Errr, wrong. You can't magically increase pressure. If the recycle has the impact of rising inlet pressure into the compressor, this will back up along the inlet system.

You can only flow from a position of higher energy to one of lower energy, in this case pressure.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[pyMorty]

Thank ou LittleInch, I see I´ve been confused about recycle valve operation,it won´t increase suction pressure to a higher value thant the original setpoint was, but just to compensate whichever pressure is missing to reach the setpoint again, as gwalkerb did mention.

 

[zdas04]

Now that that is cleared up. Can you describe your process a bit? A three stage compressor can provide something like 64 compression ratios. If you start at 600 psig that gets the pressure up to something in the range of 38,000 psig. Ariel has never made a compressor designed for 38,000 psig discharge that I've ever been able to find. What is going on here (or are you just making up numbers to clarify points)?

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[pyMorty]

zdas04, I was just making up some pressures in the last messages, since when I started the topic I was working ona Gas Gathering (three stage, target 8 MMscfd, suction around 30 psig, discharge 900 psig), where the calculated was giving 10 MMscfd; but then another project came up, a rotative closed gas lift, where the suction pressure is around 600 psig,discharge 3000 psig, 2-stage recip, i didn´t change the flow I first mentioned to avoid confusions, but I see it caused the opposite, sorry about that. What happens in the latter is as follows:

-The 2-phase separator is taking the production from the well at 700 psig, currently separating 3 MMscfd, and feeding two recips, the target injection flow for one of the wells is 1.2 MMscfd, the calculated one is 1.6 MMscfd at those conditions, (the Index of Productivity shows the best performance at 1.2 MMscfd). We would normally control de pressure at the separator by means of a so called "Excess gas valve" decreasing the pressure that feeds the compressor, in this case we would have gotten it down to 600 psig aprox., but as it is feeding a second compressor that needs the 700 psig to meets its flow (around 1.8 MMscfd) we can´t control the flow for compressor 1 with that valve

- The operators on site have slowed down the 1800 RPM nat gas engine to 1240 RPM (which is quite lower than it would be recommended) in order to get the flow down to 1.2 MMscfd. The flow computer now reads 1.2 MMscfd, but they think it is not working properly. I told them it would be better to speed the engine up to at least 1400 RPM and openning the recycle a little bit more, but they didn´t want to speed it up because it would increase the flow and were not sure about the valve´s oppening required, I told him with the Ariel Software I can only check what it is teorically compressing at 700 psig, without the effect of an openned recycle valve, which is 1.6 MMscfd. Then I got confused on whether those 700 psig were from the separation itself or the effect of having the rec valve openned (which has been already clarified it is not the way it works). The control loop is between the recycle valve and the flow computer at suction, in this case would the control loop need to be made between the flow computer at discharge and the recycle valve? Since now I see two perspectives for the Recycle valve function, one for compensating decreasing suction pressure, and the other for decreasing injection when all the other control capacity devices are not enough (this unit has no VVP´s). I will check on the valve´s Cv.